RNA-seq in the tetraploid Xenopus laevis enables genome-wide insight in a classic developmental biology model organism

Abstract

Advances in sequencing technology have significantly advanced the landscape of developmental biology research. The dissection of genetic networks in model and non-model organisms has been greatly enhanced with high-throughput sequencing technologies. RNA-seq has revolutionized the ability to perform developmental biology research in organisms without a published genome sequence. Here, we describe a protocol for developmental biologists to perform RNA-seq on dissected tissue or whole embryos. We start with the isolation of RNA and generation of sequencing libraries. We further show how to interpret and analyze the large amount of sequencing data that is generated in RNA-seq. We explore the abilities to examine differential expression, gene duplication, transcript assembly, alternative splicing and SNP discovery. For the purposes of this article, we use Xenopus laevis as the model organism to discuss uses of RNA-seq in an organism without a fully annotated genome sequence.

Keywords: Differential expression; RNA-seq; Xenopus.

Figure 1. Fragment analysis of RNA-seq libraries

Bioanalyzer trace files for final RNA-seq libraries prepared from stage 37 X. laevis hearts. Graph represents DNA content with amount on y-axis measured by fluorescent unit (FU) and fragment size on the x-axis. Two peaks at 35 and 10380 base pair (bp) correspond to DNA standards used. Note the abundance of DNA in the 300–600 bp range.

Figure 2. RNA-seq data interpretation and quality analysis

Overview of fastq file format and example of a single read output for wild type heart sample.

Figure 3. Output from fastqc software for RNA-seq data quality analysis

(A) Quality score per base position for reads. x-axis represents base pair position and y-axis represents interquartile range of quality value (from 0 to 40). Note the overall reduction in quality of scores toward the end of the 76 bp reads. (B) Graphical summary of quality scores shown in (A). (C) G/C distribution over all sequence reads represented on x-axis, with total reads corresponding on y-axis. Overall distribution (red line) versus the theoretical distribution in sample (blue line).

Figure 4. Determining morpholino (MO) efficacy by RNA-seq

(A) RT-PCR of cDNA generated from wild type and casz1 MO hearts shows aberrant splicing of casz1 intron 8 in MO-treated embryos. (B) Modified IGV viewer screenshot with schematic of casz1 mRNA (boxes) with inclusion of intron 8. Arrows indicate primers used for PCR shown in (A). Overall read coverage and individual reads mapping to casz1 mRNA in wild type and casz1 MO hearts are shown below schematic. Exon-intron boundaries predicted by the program Tophat based on read coverage are displayed on genome browser.